VerChol -- Grammar-First Tokenization for Agglutinative Languages
This addresses a critical bottleneck in LLM pipelines for speakers and developers of agglutinative languages, offering a domain-specific solution.
The paper tackles the problem of tokenization for agglutinative languages, where existing methods like Byte Pair Encoding fragment words and increase token counts, by proposing a grammar-first approach that improves efficiency and accuracy.
Tokenization is the foundational step in all large language model (LLM) pipelines, yet the dominant approach Byte Pair Encoding (BPE) and its variants is inherently script agnostic and optimized for English like morphology. For agglutinative languages a typological class encompassing the Dravidian family (Tamil, Kannada, Telugu, Malayalam), Turkic languages (Turkish, Azerbaijani, Uzbek), Uralic languages (Finnish, Hungarian, Estonian), Korean, Japanese, Swahili, Basque, and others, a single word may encode root, tense, aspect, person, number, gender agreement, case, and postpositions into one orthographic unit. Statistical tokenizers fragment these words into byte pair chunks that sever morpheme boundaries and inflate token counts.